Maize is the third most important cereal crop after wheat and rice and it is grown on about 120 million hectares worldwide. This crop is a major source of food and feed but, unfortunately, it is also a well-known host for toxigenic fungi as Aspergillus flavus able to contaminate the ripening kernels with aflatoxin B1(AFB1) which is reported as carcinogenic for humans and animals. EU legislation fixed AFB1 thresholds for raw maize destined to humans, dairy animals and other animal species (Commission Regulation 1181/2006 and Directive 100/2003). Modelling of the interactions between host, plant and environment during the season can enable prediction of pre-harvest AFB1 risk and its potential management. In this work the relational diagram of A. flavus infection cycle was developed; state variables, rates and driving variables were decided and organized in a coherent structure. Quantitative data for crucial steps of the cycle were collected from literature and equations were elaborated to connect driving variables to rates; an algorithm was then developed to finalize the model. The model predicts the risk of maize contamination by AFB1 above the legal limit of 5μg/kg. The model was validated with a six year data set and around 70% of maize samples were correctly classified, below or above the threshold, by AFLA-maize. Therefore, AFLA-maize, giving a prediction on a daily base, allows following the risk dynamic along the season and it is a useful support to alert farmers and technicians. Apart real time predictions, historical and predicted data can be used as input to draw risk maps in poorly studied areas or in climate change scenarios.